Computes exp(x) - 1
element-wise.
tf.math.expm1( x, name=None )
i.e. exp(x) - 1
or e^(x) - 1
, where x
is the input tensor. e
denotes Euler's number and is approximately equal to 2.718281.
x = tf.constant(2.0) tf.math.expm1(x) ==> 6.389056 x = tf.constant([2.0, 8.0]) tf.math.expm1(x) ==> array([6.389056, 2979.958], dtype=float32) x = tf.constant(1 + 1j) tf.math.expm1(x) ==> (0.46869393991588515+2.2873552871788423j)
Args | |
---|---|
x | A Tensor . Must be one of the following types: bfloat16 , half , float32 , float64 , complex64 , complex128 . |
name | A name for the operation (optional). |
Returns | |
---|---|
A Tensor . Has the same type as x . |
© 2020 The TensorFlow Authors. All rights reserved.
Licensed under the Creative Commons Attribution License 3.0.
Code samples licensed under the Apache 2.0 License.
https://www.tensorflow.org/versions/r2.4/api_docs/python/tf/math/expm1